Deploy AtomicCD On Kubernetes Cluster

This tutorial shows how to deploy AtomicCD a Kubernetes Cluster with a Track Config file placed on a public Github repository.

Pre-requisites:

• A Kubernetes cluster with admin privileges.
• A Github public repository to place Target Config file.
• If repository is private: A Personal Access Token of the private repository (with read access for the repository contents). If using Github Fined-grained Tokens, the Contents permission with Read-only access can be used.

Below are the steps to setup AtomicCD to implement continuous delivery for a python application:

1. Create Cluster Role for AtomicCD:

   • Make a file named cluster-role.yaml and copy the contents of the below cluster role file in it.

       apiVersion: rbac.authorization.k8s.io/v1
       kind: ClusterRole
       metadata:
         name: sa-role
       rules:
       - apiGroups: ["apps"]
         resources: ["deployments", "statefulsets"]
         verbs: ["list", "patch"]
       - apiGroups: [""]
         resources: ["pods"]
         verbs: ["list"]
   • Apply the file using kubectl apply -f cluster-role.yaml command.

2. Create a Service Account for AtomicCD

   • Make a file named sa.yaml and copy the contents of the below service account file in it.

   apiVersion: v1
   kind: ServiceAccount
   metadata:
     name: sa
     namespace: default
   • Apply the file using kubectl apply -f sa.yaml command.

3. Create a Cluster Role Binding to bind the above Service Account and Cluster Role with each other.

   • Make a Cluster Role Binding file name cluster-role-binding.yaml and copy the contents of the below cluster role binding file in it.

   apiVersion: rbac.authorization.k8s.io/v1
   kind: ClusterRoleBinding
   metadata:
     name: sa-binding
   roleRef:
     apiGroup: rbac.authorization.k8s.io
     kind: ClusterRole
     name: sa-role
   subjects:
   - kind: ServiceAccount
     name: sa
     namespace: default
   • Apply the file using kubectl apply -f cluster-role-binding.yaml command.

4. In your Github repository make a target config file named targetConfig.yaml at root directory and copy the below target config files in it.

targetConfig:
  - targetName: python-cd
    containers:
    - containerName: "python-container"
      containerImage: "python"
      containerTag: "3.9"
    name: python-deployment
    type: deployment
    namespace: default
    scope: nameScoped

5. Deploy a python deployment in your Kubernetes Cluster.

   • Create a file named python-deployment.yaml and copy the python deployment manifest below in it.

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: python-deployment
   spec:
     selector:
       matchLabels:
         app: python-deployment
     template:
         metadata:
           labels:
             app: python-deployment
         spec:
           containers:
           - name: python-container
             image: python:3.9
             command: ["sh", "-c"]
             args:
             - |
                 bash -c "sleep 99999999"
             resources:
               limits:
                 memory: "128Mi"
                 cpu: "50m"
             ports:
             - containerPort: 80
   • Apply the file with kubectl apply -f python-deployment.yaml command.

6. Create a Configmap name with a Track Config file. Follow 6a if repository if public otherwise follow 6b

   • (6a) Make a file named configmap.yaml. Below in the track config configmap, replace "your github repository url" with your actual repository URL. Copy the final contents in your configmap.yaml file. Make sure the branch is main.

   apiVersion: v1
   kind: ConfigMap
   metadata:
       name: atomic-cd-cm
   data:
       trackConfig.yaml: |
           trackConfig:
             repoURL: your github repository url
             path: /targetConfig.yaml
             token: your github personal access token
   • (6b) Make a file named configmap.yaml. Below in the track config configmap, replace "your github repository url" with your actual repository URL. Change your github personal access token with the personal access token for your github private repository. Copy the final contents in your configmap.yaml file. Make sure the branch is main.

   apiVersion: v1
   kind: ConfigMap
   metadata:
       name: atomic-cd-cm
   data:
       trackConfig.yaml: |
           trackConfig:
             repoURL: your github repository url
             path: /targetConfig.yaml
   • Apply the file with kubectl apply -f configmap.yaml command.

7. Deploy AtomicCD

   • Make a file named deploy.yaml and copy the contents of the below AtomicCD deployment manifest in it.

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: atomic-cd
   spec:
     selector:
       matchLabels:
         app: atomic-cd
     template:
         metadata:
           labels:
             app: atomic-cd
         spec:
           serviceAccountName: sa
           containers:
           - name: atomic-cd
             image: iamanshul14/atomiccd:v1
             resources:
               limits:
                 memory: "50Mi"
                 cpu: "30m"
             volumeMounts:
             - name: config-vol
               mountPath: /AtomicCD/config
             ports:
               - containerPort: 8080
           volumes:
           - name: config-vol
             configMap:
               name: atomic-cd-cm
   • Apply the file using kubectl apply -f deploy.yaml command.

Trigger CD for the python application.

• Get the name of AtomicCD pod by running the kubectl get pods command. Copy the pod name with atomic-cd prefix.

• In the command: kubectl logs your-atomiccd-pod-name --follow, change "your-atomiccd-pod-name" with the AtomicCD pod name you copied above. Run the command.

• If everything goes fine you will see running output in the container log of AtomicCD. Do not close this terminal tab and keep tracking the logs. When you see the running log in AtomicCD pod it means that AtomicCD has scanned the Track Config in repository.

• Go to your Github repository and change the value of containerTag field to 3.10 and commit the change.

• AtomicCD scans for Track Config in github repo every 3 minutes, so wait for around 3 to 6 minutes and keep a track on AtomicCD container logs. The wait time can vary depending on the time you have commited changed container tag and last executed scan of AtomicCD so a little patience is required. Also Github caches the repository data so sometimes Github can provide AtomicCD outdated TrackConfig therefore for wait for around 2 to 3 scans of AtomicCD.

• You will see the log Out of sync Deployment: python-deployment with Container Name python-container and Sync Successfull!. This means AtomicCD have succefully executed CD for python app and change the python app version from 3.9 to 3.10.

• You can confirm the changed python app version by first getting the pod name from the command: kubectl get pods and copying the pod name with prefix python-deployment and then finally running the command kubectl describe pod python-deployment-pod-name (change python-deployment-pod-name with the actual podname of python-deployment).

• In the result of above command you can check the container verson in image field.

---

Edit this page**

---